MIT Art, Design and Technology University

This study introduces a new method for classifying Diabetic Retinopathy (DR) with enhanced accuracy, focusing on addressing the limitations of existing approaches. DR classification is still a difficult task due to the presence of various complex and overlapping lesions in the retinal images. The primary goal is to develop an extremely effective model that can identify the different phases of diabetic retinopathy (DR) very accurately from the fundus images, therefore reducing the misclassification rates and helping in the very early diagnosis. The focus of our inquiry is on the determination of whether the new Hierarchical Auto-Associative Inception Polynomial Transformer Convolutional Neural Network with Snake Optimizer (HAutoAIPTCNNet + SO) will result in a significantly higher level of accuracy when compared with the existing methods. Fundus images from the EyePACS and DIARETDB1 datasets are processed using Gradient Domain Guided Filtering (GDGF) for enhanced contrast, noise reduction, and normalization. Segmentation of DR-affected regions is achieved with the EfficientNet and Cascaded Visual Attention Network (ENet-CVAN) framework. The classification process then employs the Hierarchical Auto-Associative Inception Polynomial Transformer Convolutional Neural Network (HAutoAIPTCNNet) further refined through the Snake Optimizer (SO). The HAutoAIPTCNNet + SO model, developed in Python, features both hierarchical and polynomial transformations for the precise imaging of retinal characteristics. The suggested HAutoAIPTCNNet + SO approach reached a classification accuracy of 99.8

Ovarian Cancer is a leading cause of mortality among women globally, primarily due to lack of specific and sensitive early-stage diagnostic tools. This study aims to identify hub genes associated with recurrent, late-stage, and metastatic tumors as potential prognostic biomarkers and drug targets. Gene expression data from eight National Center for Biotechnology Information (NCBI)-Gene Expression Omnibus (GEO) datasets were categorized by recurrence, tumor-stage, and metastasis. Differential gene expression and enrichment analyses were performed. Hub genes were identified by protein-protein interaction networks and validated by the University of Alabama at Birmingham Cancer Data Analysis Portal (UALCAN), GEPIA2, pROC, and Kaplan-Meier plotter databases. Genetic alterations, immune cell infiltration, miRNA prediction, and drug-gene interactions were assessed using cBioPortal, CIBERSORTx, Encyclopedia of RNA Interactomes (ENCORI), and Drug-Gene Interaction Database (DGIdb), respectively. Eight hub genes (FN1, COL1 A1, COL1A2, COL3A1, POSTN, LUM, IGF1, and CXCL8) were identified, with COL1A2 common across all tumor categories. Note that 19.6% of cases showed mutations in these genes, primarily COL3A1. Overexpression of most hub genes and reduced expression of CXCL8 correlated with worse survival outcomes. COL1A1 and FN1 showed strong diagnostic ability. Late-stage tumors showed elevated M2 macrophages and neutrophils. hsa-miR-29a-3p, hsa-miR-29b-3p, and hsa-miR-29c-3p were identified as the most interactive miRNAs. Ocriplasmin and pamidronate were identified as potential therapeutics. Our findings highlight the therapeutic relevance of these hub genes and identify them as potential drug targets and prognostic biomarkers in ovarian cancer.

Acute myeloid leukemia (AML) is a rapidly progressing blood cancer with poor survival rates, necessitating aggressive treatment strategies like chemotherapy. Doxorubicin (DOXO) is commonly used but is limited by severe side effects, including myeloablation, which involves the depletion of bone marrow cells leading to immunosuppression and heightened infection risk. This study explores the potential of omega-3 polyunsaturated fatty acids (n-3 PUFAs), specifically docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA), to enhance the efficacy of DOXO against AML cells while mitigating some of its toxicities. The results show that DHA and EPA increase the DOXO-induced apoptosis in KG1a cells and greater accumulation in the sub-G1 phase, suggesting enhanced cell death. TUNEL assays confirmed increased DNA fragmentation, whereas mRNA analysis revealed upregulation of apoptosis and cell cycle regulation genes. Importantly, DHA and EPA also reduced the hemolytic activity of DOXO, suggesting a protective effect against chemotherapy-associated side effects. These findings suggest that DHA and EPA could enhance the anti-leukemic impact of DOXO, potentially reducing the need for high-dose chemotherapy and alleviating risks like myeloablation, offering a promising adjunct strategy for AML treatment.

The advent of drone technology has led to groundbreaking advancements across various industries, including warehousing operations. In recent years, warehouse drones have garnered significant attention due to their potential to revolutionize traditional inventory management and order fulfillment processes. This paper presents a comprehensive review that synthesizes findings from more than 120 research papers on drone-enabled practices in warehouses. The review systematically considers multiple parameters, including drone function (inventory counting, mapping, surveillance, inspection, and intralogistics support), robot platforms used (UAV, UAV-AGV), deployment architecture (single and multi-drone system), validation approach (real-time and simulation), technology and methodology used (modern electronic devices, AI, and IOT), and environmental context (dynamic and static). Furthermore, the paper explores the diverse applications of warehouse drones in inventory management, maintenance and inspection, picking and packaging, goods transportation, security and surveillance, and warehouse layout optimization. The review highlights that most studies still rely on single-UAV systems tested mainly in simulations, with only a few real-time demonstrations of fully autonomous performance inside real warehouses. Although multi-drone approaches are emerging to improve scalability, they continue to struggle with coordination and safety. Research remains largely focused on static environments, with dynamic warehouse conditions receiving far less attention despite their practical importance. The findings of the review are presented with the tabulated results and a comparative table to provide a better understanding of the review work, which helps to identify the existing literature gap. The review presents its findings through clear tables and comparisons, making it easier to understand existing studies and pinpoint the gaps in the current literature.

IntroductionEarly diagnosis of endometrial cancer is critical for improving survival. However, most AI-based methods are developed using single-modality data, and most of them are uninterpretable and do not offer privacy protection, or they completely ignore privacy issues. We propose a multimodal AI framework that utilizes histopathology whole-slide images (WSIs) and clinical data and incorporates both explainability and privacy-aware learning.MethodsFive hundred and twenty-nine patients’ data (354 early-stage and 175 advanced-stage) with 794 WSIs and 208,000 image patches were collected for the study. By using a convolutional neural network (CNN), morphological features were extracted from WSIs, and clinical variables were encoded by a multi-layer perceptron, respectively. These two modalities of information (the learned representations) were combined to make the final category prediction. Interpretability was enabled through Grad-CAM and clinical feature attribution, and privacy-aware training was supported by secure parameter aggregation.ResultsThe multimodal model obtained an accuracy of 0.91 and an AUC of 0.95, thus it exceeded the performance of clinical-only (accuracy = 0.78, AUC = 0.81) and histopathology-only (accuracy = 0.85, AUC = 0.89) models with a considerable increase in sensitivity (0.89) and specificity (0.93). The performance was preserved by privacy-aware learning as well. The net clinical benefit was maximum as per the Decision Curve Analysis.ConclusionThis framework offers a solution that is accurate, interpretable, and privacy-preserving, thus it can act as a diagnostic aid for early endometrial cancer. However, since the model was developed and evaluated using only the TCGA-UCEC cohort, external multi-center validation is required to confirm generalizability across diverse clinical populations, imaging protocols, and laboratory conditions.